A Fast Algorithm for Operational Rolling Stock Assignment under Daily and Monthly Maintenance Constraints

Abstract

In the process of rail transport scheduling (or planning), we focus on an operational rolling stock rostering problem. The objective is to minimize the number of executed maintenance, and the primary constraints are to cover all train trips for the next several days by rolling stock units and to satisfy daily and monthly maintenance constraints on them. However, no study in the literature has optimally solved a practical instance with such maintenance constraint. This paper provides a framework to solve a practical instance optimally in reasonable time by adopting a branch-and-price (or column generation combined with branch-and-bound) approach and by devising an efficient method to answer the column generation subproblem which appears repeatedly in branch-and-price.